The present invention relates generally to implantable endoluminal medical devices which contact biological fluids and tissues on at least one surface of the medical device. More particularly, the present invention relates to endoluminal stents which are implanted into anatomical passageways using minimally invasive techniques. Endoluminal stents are frequently used post-angioplasty in order to provide a structural support for a blood vessel and reduce the incidence of restenosis following percutaneous balloon angioplasty. A principal example of the present invention are endovascular stents which are introduced to a site of disease or trauma within the body's vasculature from an introductory location remote from the disease or trauma site using an introductory catheter, passed through the vasculature communicating between the remote introductory location and the disease or trauma site, and released from the introductory catheter at the disease or trauma site to maintain patentency of the blood vessel at the site of disease or trauma.st endol
While the use of endoluminal stents has successfully decreased the rate of restenosis in angioplasty patients, it has been found that a significant restenosis rate continues to exist even with the use of endoluminal stents. It is generally believed that the post-stenting restenosis rate is due, in major part, to a failure of the endothelial layer to regrow over the stent and the incidence of smooth muscle cell-related neo-intimal growth on the luminal surfaces of the stent. Injury to the endothelium, the natural nonthrombogenic lining of the arterial lumen, is a significant factor contributing to restenosis at the situs of a stent. Endothelial loss exposes thrombogenic arterial wall proteins, which, along with the generally thrombogenic nature of many prosthetic materials, such as stainless steel, titanium, tantalum, Nitinol, etc. customarily used in manufacturing stents, initiates platelet deposition and activation of the coagulation cascade, which results in thrombus formation, ranging from partial covering of the luminal surface of the stent to an occlusive thrombus. Additionally, endothelial loss at the site of the stent has been implicated in the development of neointimal hyperplasia at the stent situs. Accordingly, rapid re-endothelialization of the arterial wall with concomitant endothelialization of the body fluid or blood contacting surfaces of the implanted device, is considered critical for maintaining vasculature patency and preventing low-flow thrombosis.
At present, most endoluminal stents are manufactured of stainless steel, which is known to be thrombogenic. In order to reduce the thrombogenicity of the stainless steel and to maintain sufficient dimensional profiles for catheter delivery, most stents minimize the metal surface area which contacts blood, in order to minimize thrombus formation after implantation. Thus, in order to reduce the thrombogenic response to stent implantation, as well as reduce the formation of neointimal hyperplasia, it would be advantageous to increase the rate at which endothelial cells form endothelium proximal and distal to the stent situs, migrate onto and the endothelial coverage of the luminal surface of the stent which is in contact with blood flow through the vasculature.
The surface of a solid, homogeneous material can be conceptualized as having unsaturated inter-atomic and intermolecular bonds forming a reactive plane ready to interact with the environment. In practice, a perfectly clean surface is unattainable because of immediate adsorption of airborne species, upon exposure to ambient air, of O, O2, CO2, SO2, NO, hydrocarbons and other more complex reactive molecules. Reaction with oxygen implies the formation of oxides on a metal surface, a self-limiting process, known as passivation. An oxidized surface is also reactive with air, by adsorbing simple, organic airborne compounds. Assuming the existence of bulk material of homogeneous subsurface and surface composition, oxygen and hydrocarbons may adsorb homogeneously. Therefore, further exposure to another environment, such as the vascular compartment, may be followed by a uniform biological response.
Current metallic vascular devices, such as stents, are made from bulk metals made by conventional methods, and stent precursors, such as hypotubes, are made with many steps each of which introduce processing aides to the metals. For example, olefins trapped by cold drawing and transformed into carbides or elemental carbon deposit by heat treatment, typically yield large carbon rich areas in 316L stainless steel tubing manufactured by cold drawing process. The conventional stents have marked surface and subsurface heterogeneity resulting from manufacturing processes (friction material transfer from tooling, inclusion of lubricants, chemical segregation from heat treatments). This results in formation of surface and subsurface inclusions with chemical composition and, therefore, reactivity different from the bulk material. Oxidation, organic contamination, water and electrolytic interaction, protein adsorption and cellular interaction may, therefore, be altered on the surface of such inclusion spots. Unpredictable distribution of inclusions such as those mentioned above provide an unpredictable and uncontrolled heterogeneous surface available for interaction with plasma proteins and cells. Specifically, these inclusions interrupt the regular distribution pattern of surface free energy and electrostatic charges on the metal surface that determine the nature and extent of plasma protein interaction. Plasma proteins deposit nonspecifically on surfaces according to their relative affinity for polar or non-polar areas and their concentration in blood. A replacement process known as the Vroman effect, Vroman L. The importance of surfaces in contact phase reactions, Seminars of Thrombosis and Hemostasis 1987; 13(1):79-85, determines a time-dependent sequential replacement of predominant proteins at an artificial surface, starting with albumin, following with IgG, fibrinogen and ending with high molecular weight kininogen. Despite this variability, some of the adsorbed proteins have receptors available for cell attachment and therefore constitute adhesive sites. Examples are: fibrinogen glycoprotein receptor IIbIIIa for platelets and fibronectin RGD sequence for many blood activated cells. Since the coverage of an artificial surface with endothelial cells is a favorable end-point in the healing process, to favor endothelialization is desirable in implantable vascular device manufacture.
Normally, endothelial cells (EC) migrate and proliferate to cover denuded areas until confluence is achieved. Migration, quantitatively more important than proliferation, proceeds under normal blood flow roughly at a rate of 25 μm/hr or 2.5 times the diameter of an EC, which is nominally 10 μm. EC migrate by a rolling motion of the cell membrane, coordinated by a complex system of intracellular filaments attached to clusters of cell membrane attachment, integrin receptors, specifically focal contact points. The integrins within the focal contact sites are expressed according to complex signaling mechanisms and eventually couple to specific amino acid sequences in substrate adhesion molecules (such as RGD, mentioned above). An EC has roughly 16-22% of its cell surface represented by integrin clusters Davies P. F., Robotewskyi A., Griem M. L. Endothelial cell adhesion in real time. J. Clin. Invest. 1993; 91:2640-2652, Davies, P. F., Robotewski, A., Griem, M. L., Qualitiative studies of endothelial cell adhesion, J. Clin. Invest. 1994; 93:2031-2038. This is a dynamic process, which implies more than 50% remodeling in 30 minutes. The focal adhesion contacts vary in size and distribution, but 80% of them measure less than 6 □m2, with the majority of them being about 1 □m2, and tend to elongate in the direction of flow and concentrate at leading edges of the cell. Although the process of recognition and signaling to determine specific attachment receptor response to attachment sites is incompletely understood, regular availability of attachment sites, more likely than not, would favorably influence attachment and migration. Irregular or unpredictable distribution of attachment sites, that might occur as a result of various inclusions, with spacing equal or smaller to one whole cell length, is likely to determine alternating hostile and favorable attachment conditions along the path of a migrating cell. These conditions may vary from optimal attachment force and migration speed to insufficient holding strength to sustain attachment, resulting in cell slough under arterial flow conditions. Due to present manufacturing processes, current implantable vascular devices exhibit such variability in surface composition as determined by surface sensitive techniques such as atomic force microscopy, X-ray photoelectron spectroscopy and time of flight secondary ion-mass spectroscopy.
There have been numerous attempts to increase endothelialization of implanted stents, including covering the stent with a polymeric material (U.S. Pat. No. 5,897,911), imparting a diamond-like carbon coating onto the stent (U.S. Pat. No. 5,725,573), covalently binding hydrophobic moieties to a heparin molecule (U.S. Pat. No. 5,955,588), coating a stent with a layer of blue to black zirconium oxide or zirconium nitride (U.S. Pat. No. 5,649,951), coating a stent with a layer of turbostratic carbon (U.S. Pat. No. 5,387,247), coating the tissue-contacting surface of a stent with a thin layer of a Group VB metal (U.S. Pat. No. 5,607,463), imparting a porous coating of titanium or of a titanium alloy, such as Ti—Nb—Zr alloy, onto the surface of a stent (U.S. Pat. No. 5,690,670), coating the stent, under ultrasonic conditions, with a synthetic or biological, active or inactive agent, such as heparin, endothelium derived growth factor, vascular growth factors, silicone, polyurethane, or polytetrafluoroethylene, U.S. Pat. No. 5,891,507), coating a stent with a silane compound with vinyl functionality, then forming a graft polymer by polymerization with the vinyl groups of the silane compound (U.S. Pat. No. 5,782,908), grafting monomers, oligomers or polymers onto the surface of a stent using infrared radiation, microwave radiation or high voltage polymerization to impart the property of the monomer, oligomer or polymer to the stent (U.S. Pat. No. 5,932,299). Thus, the problems of thrombogenicity and re-endothelialization associated with stents have been addressed by the art in various manners which cover the stent with either a biologically active or an inactive covering which is less thrombogenic than the stent material and/or which has an increase capacity for promoting re-endothelialization of the stent situs. These solutions, however, all require the use of existing stents as substrates for surface derivatization or modification, and each of the solutions result in a biased or laminate structure built upon the stent substrate. These prior art coated stents are susceptible to delamination and/or cracking of the coating when mechanical stresses of transluminal catheter delivery and/or radial expansion in vivo. Moreover, because these prior art stents employ coatings applied to stents fabricated in accordance with conventional stent formation techniques, e.g., cold-forming metals, the underlying stent substrate is characterized by uncontrolled heterogeneities on the surface thereof. Thus, coatings merely are laid upon the heterogeneous stent surface, and inherently conform to the heterogeneities in the stent surface and mirror these heterogeneities at the blood contact surface of the resulting coating. This is conceptually similar to adding a coat of fresh paint over an old coating of blistered paint, the fresh coating will conform to the blistering and eventually, itself, blister and delaminate from the underlying substrate.
The current invention entails creating materials specifically designed for manufacture of stents and other intravascular devices. Manufacture of stents and other intravascular devices is controlled to attain a regular, homogeneous atomic and molecular pattern of distribution along their surface. This avoids the marked variations in surface composition, would create a predictable oxidation and organic adsorption pattern and would have a predictable interaction with water, electrolytes, proteins and cells. Particularly, EC migration would be supported by a homogeneous distribution of binding domains which serve as natural or implanted cell attachment sites, in order to promote unimpeded migration and attachment. Based on observed EC attachment mechanisms such binding domains should have a repeating pattern along the blood contact surface of no less than 1 micron radius and 2 micron border to border spacing between binding domains. Ideally, the inter-binding domain spacing is less than the nominal diameter of an endothelial cell in order to ensure that at any given time, a portion of an endothelial cell is in proximity to a binding domain.